Aspirations to harness the lethal potency of plant and bacterial toxins for beneficial use in medicine have yet to be fully realized. Various toxins have been chemically or genetically linked with specific antibody and ligand carriers to focus their action exclusively on cancer cells. These new derivative toxins provide very potent and highly selective cell kill in vitro but their performance as therapeutic agents in animal models has continued to fall far short of expectation. Substantial deficiencies have persisted despite years of intensive effort to refine the technology for producing toxin conjugates. This continuing study has therefore been designed to identify the key factors underlying in vivo effectiveness and to furnish unique insights into tumoricidal mechanisms. It addresses this critical problem from a novel perspective by deploying a model system of advanced neoplastic disease which is curable using selective toxin therapy. Valuable information regarding the fundamentals of toxin-based therapeutics is being obtained which should facilitate the design of agents with optimal clinical effectiveness and utility. A model of human malignant mesothelioma in athymic and SCID mice is being used, since the natural resistance of murine cells allows diphtheria toxin to selectively kill the human cancer target cells and cure these mice. Thus this study will encompass patterns of malignancy which involve mainly the peritoneal cavity as well as more disseminated forms of disease. Special in vivo tumor labeling and autoradiographic techniques have been developed to measure the time course, location and extent of native toxin versus immunotoxin action on tumors in situ. Essential features are thereby being revealed which explain how a single microgram dose of native toxin can kill a billion or more cancer cells and eradicate established solid tumors weighing 1-3 gm. A new, diphtheria toxin-based hybridoma screening agent has been designed to identify the most suitable monoclonal antibodies for making potent anti-tumor cell immunotoxins. Experiments are being performed to determine the pertinent distinctions which explain discrepancies in the action of toxin conjugates and diphtheria toxin in this model. Various antibodies, antibody fragments and ligands will be linked to diphtheria toxin to determine if such modifications curtail its access to tumor cells and diminish its effectiveness. The therapeutic model will provide a frame of reference to judge the performance of modified toxins and to improve those which fail to achieve cures. This system offers a means for directly testing which structural changes obstruct the curative properties of a toxin and which are permissible.